121–130 of 159 posts
http://news.ycombinator.com/item?id=1218951 http://news.ycombinator.com/item?id=478632 Further, I hope JoshTriplett (http://news.ycombinator.com/user?id=JoshTriplett) isn't too disappointed that when he submitted this exact same item 2 days ago it got one upvote and no discussion. In submitting to HN, as with comedy, timing is everything. http://news.ycombinator.com/item?id=3784912
Earlier quoted context omitted.
>What human can understand it? Lets start with Wilson's midpoint, since that's just high school math. def mid(upvotes:Int, downvotes:Int) = { val total = upvotes+downvotes+0.0 val up = upvotes/total val half = 0.5 val a = total/(4+total) val b = 4/(4+total) a * up + b * half } So there are two weights a and b. Using these weights, the midpoint is a weighted average of half and the proportion of upvotes. It should be…
The problem isn't that it's impossible for humans to understand, nor that one can't hire the best in the field to implement this solution. The problem is that the people who need to understand the outcome won't be able to understand it. Case in point, the article mentions Amazon. Amazon could easily hire the best and brightest mathematicians and developers to make the implementation completely rigorous and provably c…
This has been the number one complaint I have against Amazon for the past 10 years. And they haven't done a think about it?
I love it and I hate it. Why I love it: It's precise. It's elegant. It's rigorous. It's based upon solid, proven science & theory. It's a perfect application for a computer. And most of all, it does what's intended: it works. Why I hate it: What human can understand it? I used to implement the first manufacturing and distribution systems that used thinking like this. They figured, "We finally have the horsepower to a…
That's why a friend of mine joined the army engineers. As a civil engineer working for a local city he might be involved in a 10year process of approvals to add a freeway on ramp. Where most of his job would be checking that an army of subcontractors were all doing things to code - not that they were doing things well, just to the written requirements In Afghanistan if they want a road or a barrier he basically finds…
Original author here. For the academically inclined, there is a critique of this approach in this paper: http://www.dcs.bbk.ac.uk/~dell/publications/dellzhang_ictir2... Of course, I think the authors miss the point of the algorithm, since I basically wanted a system that is one-sided (i.e. false negatives are OK but false positives are bad). Also, if you deal with more than two outcomes you might be interested in mul…
I appreciate people who take the time to apply math to things in the real world, and share it with non academic crowds. Thanks for that. 5 star rating systems are obnoxious. From a mathematical perspective, if you treat them in an ordinal fashion they are poorly behaved, and if you treat them categorically, you lose the relationship between stars. There seems to be some popular movement towards binary rating systems,…
Helping out a friend with a statistics test, I recently read up about the Wilcoxon Signed Rank Test[1]. Now this one is intended to get a p-value for experiments with "before" and "after" measurements, but what I got the idea it's trying to do, is to use the ranks of a not-very-normal behaving random variable, turn that into a summation of lots of things, so due to the central limit theorem you can treat it as a normal distribution again.
Though thinking about it, in this case it's the rank we're after, so maybe it's not useful at all. But it gives an interesting idea about the tricks you can pull if your input data isn't quite the sort of type that you can analyse very well.
Earlier quoted context omitted.
You could also go one step further on the Bayesian path and infer even alpha from the data on your site, and introduce a loss function on your ordering. Or you could do a semi-frequentist thing and simplify your math by using MAP estimates to rank. Basically instead of score = #pos/(#pos + #neg), it becomes score = (#pos+x)/(#pos+x + #neg+y), where you choose x and y to suit your needs. You could choose x/y in propor…
I read the article, went "The ^&#$?" and came here to post this, more or less. It's amazing the $&#! people will get up to when they don't know Bayesian statistics.
Original author here. For the academically inclined, there is a critique of this approach in this paper: http://www.dcs.bbk.ac.uk/~dell/publications/dellzhang_ictir2... Of course, I think the authors miss the point of the algorithm, since I basically wanted a system that is one-sided (i.e. false negatives are OK but false positives are bad). Also, if you deal with more than two outcomes you might be interested in mul…
I really appreciated the post. Unfortunately we're using 5 stars, and need to do it a bit less wrong. The main thorniness of 5 stars is that you have to answer the question of what the difference in star ratings actually mean. Is going from one star to two stars the same as four to five? Probably not, based on the way users rate, which means an algorithm like the arithmetic mean that treats them the same is wrong. Pe…
Typically, you do not know that, especially if you are comparing across reviewers. Some will only score zero and 5 stars, other will have 10% two stars, 80% three stars, and 10% four stars, yet others will have 10% three stars, 80% four stars, and 10% five stars.
If you have sufficient data (rare), it may be possible to (somewhat) correct for that. IIRC, this was something that helped in winning the Netflix challenge.
For example, http://www.netflixprize.com/assets/GrandPrize2009_BPC_Pragma... models the assignment of stars as two parts:
- modeling the user's appreciation of the movie - modeling how the user translates his appreciation to a star rating
"where we know that a higher rating is better but assume nothing beyond that"
Problem with that is that you throw away information with that assumption. You do know that 2 star scores are very unlikely to be about very good items.
Earlier quoted context omitted.
I think the original post still stands - try explaining the above to an irate customer.
I've dealt with people who have wanted a retraction of negative comments to their business on the web, I explained it didn't work that way as best I could, but there isn't much you can do. Same with this, if they are irate about a low rating, you play nice, but you don't expect them to learn the ins and outs.